Impactr starts from one observation: the value of sustainability is known and the risks are real, but knowing how to navigate both is not. Most of that gap isn't economics. It's the cost of working out the economics. Everything we build follows from that.
The business case is no longer in doubt: sustainability interventions return an estimated 2–14× over their lifetime, companies with strong sustainability practice carry measurably higher valuations, and five of the ten most severe long-term global risks are now environmental. It is no longer a question of whether to act. Only what to do, in what order, and at what cost.
So why doesn't it happen? Because every decision carries a cost before a euro is spent on the action itself: researching options, modelling trade-offs, finding grants, coordinating across teams. The constraint isn't the economics of the interventions. It's the cost of getting to them.
The gross benefit (lower emissions, energy, water and waste, plus avoided regulatory and physical risk) and the gross cost (equipment, contracts, staff time) are both broadly knowable. It's the third term, the transaction cost of deciding, that quietly kills viable actions before they're ever properly evaluated. That hidden barrier is what Impactr removes.
AI has collapsed the cost of analysis. It can't change an intervention's technology or payback, but it removes the transaction cost of finding and evaluating it. When the cost of deciding falls, the bar for action falls with it: actions never worth the effort to assess now clearly clear a positive return. Far more of the available value becomes reachable.
There's a fair objection to face here: AI consumes a great deal of energy, much of it in data centres, and the people most serious about sustainability are right to ask about it.
The answer is in the comparison. The relevant measure isn't AI's energy use against zero. It's AI's energy use against the sustainability action it unlocks. When AI surfaces a decision an organisation would otherwise never have reached, across energy, water, food and drink, waste, travel and transport, and the materials in its operations, the impact of that action should run well ahead of the energy spent to find it.
It's the same logic as the barrier above: a cost, weighed against the value it makes reachable. Multiply it across all the stranded sustainability actions sitting unmade across an economy, and the net effect of AI, its own energy included, can turn positive.
We don't say this to wave the concern away. AI's footprint is real, which is exactly why it should be pointed at the decisions where it unlocks disproportionate value, not spent on everything.
Speed is the easy half. A recommendation is only as good as the model behind it, and speed without judgement just reaches the wrong answer faster. The real problem to solve is trust: decisions an organisation can stand behind. That takes more than a generic chatbot: a model that reasons about sustainability the way an expert would.
The Impactr Decision Model turns AI speed into decisions you can trust: five principles behind a continuously reliable, least-cost pathway across impact and risk.